1. Problem Statement

1 billion hours of Youtube are watched per day. It is the 2nd most visited site in the world.

YouTube attracts 44% of all internet users whereas 37% of all mobile internet traffic belongs to YouTube

2.SENTIMENTS

we are now performing sentiment analysis with respect to "comment_text" feature on the above

to do this, we install an external library called textblob library (pip install textblob on conda prompt)

3. EDA FOR POSITIVE SENTIMENTS

we need positive polarity for this step. Therefore we are going to set the positive polarity as ==1

I am going to create and parse a filter in the data frame

to visualize 'comment_text' feature we pip install worldcloud for comments. it is a tool to show how important a word is in a feature

from this, we see the most common positive words as "AWESOME, BEST, VIDEO, PERFECT".

The bigger the size of the word, the bigger the polarity it has

4. EDA FOR NEGATIVE SENTIMENTS

5. ANALYSING TAGS COLUMN, WHAT ARE THE TRENDING TAGS ON YOUTUBE?

6. ANALYSIS ON LIKES, VIEWS & DISLIKES: HOW THEY CORRELATE TO EACH OTHER

I'll use a regression plot because i'm trying to find the relationship between two variables. This will be done using seaborn

using correlation matrix to see how they are correlated

7. EMOJI ANALYSIS ON COMMENTS